CDL Hummingbird
Faster Than Fast, Now Available via AWS PrivateLink
24th July 2019
Matt Houghton
Hummingbird's ability to complete complex searches from a range of sources and conduct real-time analytics and visualisations enables businesses to derive meaningful value from their data.
The solution's analytics capability allows organisations to match information and spot behavioural patterns in high volume data traffic in real time, enabling them to transform the customer journey and price products more competitively.
Architecture
Typically, an API call is made when data enrichment is required.We take that request and query Elasticsearch (or external APIs) for everything we know about that customer. We make use of shared data that CDL holds a licence for plus our clients' own data.
Once we have the data, we apply customer-specific business rules. This generates a data enrichment which is passed back to the calling application. The application can then make a decision based on the new insight on how to interact better with that consumer.
Creating Powerful Partnerships
Automation is key to ensuring we can quickly deploy the solution at scale in a consistent way while still maintaining high availability.Previously working with customers to provide secure connectivity to the Hummingbird APIs could be time consuming. For example, configuring VPNs was often a manual exercise.
AWS PrivateLink is a highly available, scalable technology that allows us to privately connect a customer VPC to our Hummingbird VPC endpoint services. This massively speeds up adoption of our services, for example the configuration of the customer account can be made very simple by providing a CloudFormation template.
Key Facts
~9 million requests a day as part of the customer journey
Vertically and horizontally scalable
Average response time of 120 milliseconds including external API calls
Our DevOps team delivers new functionality to production in 2-week sprints
AWS Well Architected review
99.99999% availability over the last three years
Terraform is used to define infrastructure as code
JAVA based microservices using Reactive for non-blocking threads
Blue/green deployments
Performance tested every two weeks for 25 million requests a day
Use Cases
Hummingbird has been proven in a number of different use cases, having been designed to be very flexible.For example, in the financial services market, retailers have implemented their own rules using real-time pricing to determine what action to take, based on factors such as the number of changes a consumer makes between quotes.
Use cases include:
Entity resolution
Single customer view
Tackling online fraud
Customer profiling
Optimising pricing
Transforming the customer journey
Commercial insight
In summary Hummingbird is a proven, ultra-fast data intelligence solution, which is able to search and analyse billions of records at sub-second speed. For further details please email AWSPrivateLink@fasterthanfast.com© 2024 Cheshire Datasystems Limited
Top Employer